Minoir changes

.gitignore

@@ -0,0 +1 @@
+.*.swp

common.lua

@@ -1,7 +1,4 @@
 
--- TODO: I don't like to have such tiny modules. Either merge into another
--- module or provide the functionality with another approach.
-
 --------------------------------------------------------------------------------
 -- Enum
 
@@ -27,13 +24,13 @@ local UNICODE_ZERO_WIDTH_CHARACTERS = {}
 for i = 128, 191 do  UNICODE_ZERO_WIDTH_CHARACTERS['\204'..string.char(i)] = true  end
 for i = 128, 175 do  UNICODE_ZERO_WIDTH_CHARACTERS['\205'..string.char(i)] = true  end
 
-
 local function iterate_utf8_chars (str)
 	-- TODO: Detect invalid codepoints.
 	return str:gmatch(UNICODE_CHAR_PATTERN)
 end
 
 local function utf8_string_length (str)
+	assert(type(str) == 'string')
 	local len = 0
 	for char in iterate_utf8_chars(str) do
 		if not UNICODE_ZERO_WIDTH_CHARACTERS[char] then
@@ -43,10 +40,43 @@ local function utf8_string_length (str)
 	return len
 end
 
+--------------------------------------------------------------------------------
+-- L utility
+
+local function width_of_strings_in_l (l, start_i, stop_i)
+    -- Argument fixing and Error Checking
+    assert(type(l) == 'table')
+
+    local start_i, stop_i = start_i or 1, stop_i or #l
+
+    assert(type(start_i) == 'number')
+    assert(type(stop_i)  == 'number')
+
+    -- Do stuff
+    local  width = 0
+    for i = start_i, stop_i do
+        local item_width = 0
+        if type(l[i]) == 'string' then
+            item_width = utf8_string_length(l[i])
+        elseif l[i].est_width then
+            item_width = l[i].est_width
+        end
+        width = width + item_width
+    end
+
+    -- Return
+    return width
+end
+
+
+
 --------------------------------------------------------------------------------
 
 return {
 	TABLE_TYPE = enum { 'EMPTY', 'SEQUENCE', 'STRING_MAP', 'PURE_MAP', 'MIXED', 'SET' },
 	DISPLAY = { HIDE = 1, SMALL = 2, INLINE = 3, EXPAND = 4 },
+
 	utf8_string_length = utf8_string_length,
+    width_of_strings_in_l = width_of_strings_in_l,
 }
+

function.lua

@@ -61,6 +61,7 @@ simplest, and move towards abstraction.
 local LIBRARY =  require((... and select('1', ...):match('.+%.') or '')..'library') or {}
 local DISPLAY =  assert(require((... and select('1', ...):match('.+%.') or '')..'common'), '[pretty]: Could not load vital library: common') . DISPLAY
 local utf8_string_length =  assert(require((... and select('1', ...):match('.+%.') or '')..'common'), '[pretty]: Could not load vital library: common') . utf8_string_length
+local width_of_strings_in_l =  assert(require((... and select('1', ...):match('.+%.') or '')..'common'), '[pretty]: Could not load vital library: common') . width_of_strings_in_l
 
 -- Constants
 
@@ -235,15 +236,6 @@ end
 --------------------------------------------------------------------------------
 -- Text handling
 
-local function width_of_strings_in_l (l, start_i, end_i)
-	-- FIXME: Copy of the one in pretty.lua
-    local  width = 0
-    for i = start_i or 1, (end_i or #l) do
-        width = width + utf8_string_length(l[i])
-    end
-    return width
-end
-
 local function add_indent_to_string (str, indent)
 	-- Indents `str` by `indent`.
 
@@ -380,3 +372,4 @@ return function (value, display, l, format_value)
 	end
     l[#l+1] = '\nend'
 end
+

pretty.lua

@@ -72,12 +72,17 @@ local ERROR_UNKNOWN_TYPE = [[
 We are attempting to cover all Lua features, so please report this bug, so we can improve.
 ]]
 
+local TERMINAL_WIDTH = 80
 local MAX_WIDTH_FOR_SINGLE_LINE_TABLE = 38
+
 if io and io.popen then
     local f = io.popen "tput cols"
     local term_width = f:read '*n'
     f:close()
---    if term_width then  MAX_WIDTH_FOR_SINGLE_LINE_TABLE = term_width * 3 / 2  end
+    if term_width then
+        TERMINAL_WIDTH = term_width
+        --MAX_WIDTH_FOR_SINGLE_LINE_TABLE = term_width * (2 / 3)
+    end
 end
 
 local KEY_TYPE_SORT_ORDER = {
@@ -101,6 +106,8 @@ local VALUE_TYPE_SORT_ORDER = {
     ['function'] = 7,
 }
 
+local COLUMN_SEPERATION = 1
+
 --------------------------------------------------------------------------------
 -- Key-value-pair Util
 
@@ -202,24 +209,7 @@ end
 -- Formatting Util
 
 local length_of_utf8_string  =  import 'common' . utf8_string_length
-
-local function width_of_strings_in_l (l, start_i, stop_i)
-
-    -- Argument fixing and Error Checking
-    assert(type(l) == 'table')
-
-    local start_i, stop_i = start_i or 1, stop_i or #l
-
-    assert(type(start_i) == 'number')
-    assert(type(start_i) == 'number')
-
-    -- Do stuff
-    local  width = 0
-    for i = start_i, stop_i do
-        width = width + ((type(l[i]) ~= 'string') and 1 or length_of_utf8_string(l[i]))
-    end
-    return width
-end
+local width_of_strings_in_l  =  import 'common' . width_of_strings_in_l
 
 local function ignore_alignment_info (l, start_i, stop_i)
 
@@ -229,7 +219,7 @@ local function ignore_alignment_info (l, start_i, stop_i)
     local start_i, stop_i = start_i or 1, stop_i or #l
 
     assert(type(start_i) == 'number')
-    assert(type(start_i) == 'number')
+    assert(type(stop_i)  == 'number')
 
     -- Do stuff
     for i = start_i, stop_i do
@@ -239,15 +229,14 @@ local function ignore_alignment_info (l, start_i, stop_i)
     end
 end
 
-local function fix_alignment (l, start_i, stop_i)
-
+local function insert_alignment_estimations (l, start_i, stop_i)
     -- Argument fixing and Error Checking
     assert(type(l) == 'table')
 
     local start_i, stop_i = start_i or 1, stop_i or #l
 
     assert(type(start_i) == 'number')
-    assert(type(start_i) == 'number')
+    assert(type(stop_i)  == 'number')
 
     -- Find maximums
     local max  =  {}
@@ -260,7 +249,28 @@ local function fix_alignment (l, start_i, stop_i)
     -- Insert the proper whitespace
     for i = start_i, stop_i do
         if type(l[i]) == 'table' and l[i][1] == 'align' then
-            l[i] = string.rep(' ', max[ l[i][2] ] - l[i][3])
+            l[i].est_width = max[ l[i][2] ] - l[i][3]
+        end
+    end
+end
+
+local function fix_alignment (l, start_i, stop_i)
+
+    -- Argument fixing and Error Checking
+    assert(type(l) == 'table')
+
+    local start_i, stop_i = start_i or 1, stop_i or #l
+
+    assert(type(start_i) == 'number')
+    assert(type(stop_i) == 'number')
+
+    -- Find whitespace to insert
+    insert_alignment_estimations(l, start_i, stop_i)
+
+    -- Insert whitespace
+    for i = start_i, stop_i do
+        if type(l[i]) == 'table' and l[i][1] == 'align' then
+            l[i] = string.rep(' ', l[i].est_width)
         end
     end
 end
@@ -301,12 +311,14 @@ local function align_into_columns (l, start_i, stop_i)
 	assert(type(start_i)	== 'number')
 	assert(type(stop_i)	== 'number')
 
+    insert_alignment_estimations(l, start_i, stop_i)
+
     -- Find columns
     local columns = nil
     for nr_items_pr_row = 10, 1, -1 do  -- TODO: Do this more intelligently.
         local column_width
         column_width, columns  =  attempt_to_align_into_columns(l, start_i, stop_i, nr_items_pr_row)
-        if column_width <= MAX_WIDTH_FOR_SINGLE_LINE_TABLE then  break  end
+        if column_width <= l.options.max_width_for_single_line_table then  break  end
     end
 
     -- Change alignment of columns
@@ -336,7 +348,7 @@ local function align_into_columns (l, start_i, stop_i)
                 local column_i = (item_nr-1)%#columns+1
                 if column_i ~= #columns then
                     local width_of_item = width_of_strings_in_l(l, start_of_item_i, i-1)
-                    l[i] = l[i][2] .. ' ' .. (' '):rep(columns[column_i]-width_of_item)
+                    l[i] = l[i][2] .. (' '):rep(COLUMN_SEPERATION+columns[column_i]-width_of_item)
                 end
             end
             start_of_item_i, item_nr  =  i + 1, item_nr + 1
@@ -345,7 +357,7 @@ local function align_into_columns (l, start_i, stop_i)
 end
 
 local function align_into_tabular_style (l, start_i, stop_i)
-	-- Adds alignment after seperators, to create nicely aligned tabular-format.
+	-- Adds alignment after separators, to create nicely aligned tabular-format.
 
     -- Argument fixing and Error Checking
 	local start_i, stop_i = start_i or 1, stop_i or #l
@@ -494,12 +506,14 @@ local function format_table (t, display, l, format_value)
 
     -- Decide for short or long table formatting.
     local table_width  =  width_of_strings_in_l(l, start_of_table_i)
-    if table_width <= MAX_WIDTH_FOR_SINGLE_LINE_TABLE then
+
+    if table_width <= l.options.max_width_for_single_line_table then
         -- Is short table: Ignore "width of key".
         l[start_of_table_i][3] = 'inline'
         ignore_alignment_info(l, start_of_table_i)
     elseif table_info.is_leaf_node then
         -- Is leaf node: Can format into columns.
+        -- Only if long or sequence.
         -- NOTE: Currently we only allow leaf-nodes to format into columns, due
         -- to issues with table alignment.
         align_into_columns(l, start_of_table_i)
@@ -555,7 +569,6 @@ local TYPE_TO_FORMAT_FUNC = {
 local function format_value (value, display, l)
     assert(type(display) == 'number' and type(l) == 'table')
     local formatting = TYPE_TO_FORMAT_FUNC[type(value)]
-    --print(value, formatting)
     if formatting then
         formatting(value, display, l, format_value)
     else
@@ -581,6 +594,8 @@ local KNOWN_OPTIONS = {
     _table_addr_comment = { type = 'boolean', default = false, debug = 'debug' }, -- TODO: Maybe automatically display table address when display = 0?
 
     indent              = { type = 'string',  default = '    ' },
+
+    max_output_width    = { type = 'number',  default = TERMINAL_WIDTH }
 }
 
 local function ensure_that_all_options_are_known (input_options)
@@ -615,10 +630,36 @@ local function ensure_that_all_options_are_known (input_options)
             output_options[option_name] = option_info.default
         end
     end
+    -- Calculate derived settings
+    output_options.max_width_for_single_line_table = MAX_WIDTH_FOR_SINGLE_LINE_TABLE -- TODO: Make dynamic
+
     -- Returns input_options
     return output_options
 end
 
+local function length_of_longest_line_in_text (text)
+    assert(type(text) == 'string')
+    local longest_line_len = text:match '([^\n]*)$' : len()
+    for line in text:gmatch '(.-)\n' do
+        longest_line_len = math.max(longest_line_len, line:len())
+    end
+    return longest_line_len
+end
+
+local function internal_warning (fmt, ...)
+    io.stderr:write('[pretty/internal]: '..string.format(fmt, ...))
+end
+
+local function assert_pretty_result (repr, options)
+    assert(type(repr) == 'string')
+    assert(type(options) == 'table')
+    -- Determine length of longest line in output
+    local max_width = length_of_longest_line_in_text(repr)
+    if max_width > options.max_output_width then
+        internal_warning('Internal assertion failed. Width of output was %i, but should be less than %i.', max_width, options.max_output_width)
+    end
+end
+
 local function pretty_format (value, options)
     -- Error checking
     local options  =  ensure_that_all_options_are_known(options or {})
@@ -635,7 +676,13 @@ local function pretty_format (value, options)
     fix_seperator_info(l, l.options.indent)
     ignore_alignment_info(l)
 
-    return table.concat(l, '')
+    -- Concat and perform last assertions.
+    local repr = table.concat(l, '')
+    assert_pretty_result(repr, options)
+
+    -- Return
+    return repr
 end
 
 return pretty_format
+

pstring.lua

@@ -109,14 +109,27 @@ local function safe_cut (str, si, ei)
 	-- Calculate
 	local cut_str = str:sub(si, ei)
 
-	-- Search for the number of backslashes just before the send of the string.
-	-- If that number is even, it's a sequence of backslashes, if not it's a
-	-- broken escape string.
+	-- Search for the number of backslashes and digits at the end of the string.
+	-- If the number of backslashes is even, it's a sequence of backslashes, if
+	-- not it's a broken escape string.
 	local start_of_backslashes, start_of_digits = cut_str:match '()\\*()%d?%d?$'
 	local nr_backslashes_before_end  =  start_of_digits - start_of_backslashes
-	if nr_backslashes_before_end % 2 == 1 then  cut_str = cut_str:sub(1, start_of_backslashes - 1)  end
+	if nr_backslashes_before_end % 2 == 1 then
+		-- Lets see if we can't shorten the escape code, to fit within the
+		-- cut limit.
+		local space_left = #cut_str - (start_of_digits - 2)
+		cut_str = cut_str:sub(1, start_of_digits - 2)
+		ei = ei - space_left
 
-	return cut_str
+		local digits, after_digits = str:match('^\\(%d?%d?%d?)()', si - 1 + start_of_digits - 1)
+
+		if space_left >= 1 + 3 - #digits:match '0*' then
+			ei = after_digits - 1
+			cut_str = cut_str .. ('\\%0'..(space_left-1)..'i'):format(digits)
+		end
+	end
+
+	return cut_str, ei
 end
 
 
@@ -149,8 +162,8 @@ local function format_concatted_string (str, _, l)
 	-- Cut strings
 	local sub_strings, str_i = {}, 1
 	repeat
-		local sub_str  =  safe_cut(str, str_i, str_i + width_without_overhead - 1)
-		str_i = str_i + #sub_str
+		local sub_str, ei  =  safe_cut(str, str_i, str_i + width_without_overhead - 1)
+		str_i = ei + 1
 		sub_strings[#sub_strings+1] = sub_str
 	until str_i >= #str
 

test/test_pretty.lua

@@ -34,6 +34,15 @@ local function format_test (t)
     end, { line = debug.getinfo(2).currentline })
 end
 
+local function format_parsable_test (t)
+    local stripped = t.text:match '^%s*(.-)%s*$'
+    return format_test {
+        name   = t.name,
+        input  = loadstring('return '..stripped)(),
+        expect = stripped
+    }
+end
+
 --------------------------------------------------------------------------------
 -- Primitive types
 
@@ -263,13 +272,24 @@ format_test {
     expect = '{\n    \'hello\', \'world\',       \'how\',\n    \'is\',    \'it\',          \'going?\',\n    \'Im\',    \'doing great\', \'thanks\',\n    \'that\',  \'was\',         \'what\',\n    \'I\'\n}',
 }
 
+format_parsable_test {
+    name = 'Column style with aligned numbers',
+    text = [[
+{
+    200, -522, 423, 516, 523, 126, 2912,
+    523, -620,   0,   0,   0,  -5,    2,
+     72,    6
+}
+]] }
+
 format_test {
 	name = 'Tabular style with strings left aligned',
     input  = {
 		{ a = 'hello',  b = 'hi'    },
-		{ a = 'hi',	b = 'hello'	}
+		{ a = 'hi',	    b = 'hello' },
+		{ a = 'hi',	    b = 'hi'    },
 	},
-    expect = '{\n    { a = \'hello\', b = \'hi\'    },\n    { a = \'hi\',    b = \'hello\' }\n}',
+    expect = '{\n    { a = \'hello\', b = \'hi\'    },\n    { a = \'hi\',    b = \'hello\' },\n    { a = \'hi\',    b = \'hi\'    }\n}',
 }
 
 format_test {
@@ -398,53 +418,6 @@ end)
 local BIG_EXAMPLE_TABLE = [[
 return {
     [0]   = 21082, [1]  = 696,
-    [2]   = 463,   [3]  = 235,
-    [4]   = 315,   [5]  = 312,
-    [6]   = 204,   [7]  = 124,
-    [8]   = 692,   [9]  = 84,
-    [10]  = 248,   [11] = 148,
-    [12]  = 108,   [13] = 109,
-    [14]  = 1019,  [15] = 1211,
-    [16]  = 470,   [17] = 73,
-    [18]  = 121,   [19] = 36,
-    [20]  = 149,   [21] = 514,
-    [22]  = 38,    [23] = 45,
-    [24]  = 353,   [25] = 27,
-    [26]  = 27,    [27] = 51,
-    [28]  = 84,    [29] = 61,
-    [30]  = 29,    [31] = 448,
-    [32]  = 2064,  [33] = 65,
-    [34]  = 34,    [35] = 20,
-    [36]  = 859,   [37] = 239,
-    [38]  = 24,    [39] = 41,
-    [40]  = 297,   [41] = 95,
-    [42]  = 43,    [43] = 30,
-    [44]  = 202,   [45] = 123,
-    [46]  = 243,   [47] = 98,
-    [48]  = 207,   [49] = 484,
-    [50]  = 31,    [51] = 59,
-    [52]  = 51,    [53] = 118,
-    [54]  = 27,    [55] = 22,
-    [56]  = 227,   [57] = 168,
-    [58]  = 55,    [59] = 38,
-    [60]  = 74,    [61] = 106,
-    [62]  = 62,    [63] = 40,
-    [64]  = 170,   [65] = 857,
-    [66]  = 412,   [67] = 136,
-    [68]  = 737,   [69] = 238,
-    [70]  = 64,    [71] = 119,
-    [72]  = 2567,  [73] = 481,
-    [74]  = 50,    [75] = 55,
-    [76]  = 714,   [77] = 189,
-    [78]  = 61,    [79] = 55,
-    [80]  = 114,   [81] = 26,
-    [82]  = 69,    [83] = 150,
-    [84]  = 238,   [85] = 172,
-    [86]  = 65,    [87] = 81,
-    [88]  = 102,   [89] = 39,
-    [90]  = 30,    [91] = 154,
-    [92]  = 155,   [93] = 191,
-    [94]  = 75,    [95] = 185,
     [96]  = 62,    [97] = 334,
     [98]  = 119,   [99] = 217,
     [100] = 261
@@ -458,3 +431,4 @@ end)
 --------------------------------------------------------------------------------
 
 return SUITE
+

test/test_pstring.lua

@@ -130,6 +130,13 @@ format_test {
     expect = '{\n    \'Lorem ipsum dolor sit amet, consec\\t\'...\n}',
 }
 
+format_test {
+    name   = 'Cut strings are not cut in the middle of backslash sequence',
+    not_idempotent = true,
+    input  = {'Lorem ipsum dolor sit amet, consec\\\\\\\\\\\\\\tetur adipiscing elit. Nunc vestibulum tempus ligula. Sed ac lobortis mi.'},
+    expect = '{\n    \'Lorem ipsum dolor sit amet, consec\\\\\'...\n}',
+}
+
 format_test {
     name   = 'Cut strings are not cut in the middle of escaping \\',
     not_idempotent = true,
@@ -145,13 +152,19 @@ format_test {
 }
 
 format_test {
-    -- NOTE: Not priority functionallity.
-    name   = 'Cut strings can shorten decimal escape codes, if nessesary and possible',
+    name   = 'Cut strings can shorten decimal escape codes, if necessary and possible',
     not_idempotent = true,
     input  = {'Lorem ipsum dolor sit amet, consec\014tetur adipiscing elit. Nunc vestibulum tempus ligula. Sed ac lobortis mi.'},
     expect = '{\n    \'Lorem ipsum dolor sit amet, consec\\14\'...\n}',
 }
 
+format_test {
+    name   = 'Cut strings can shorten decimal escape codes, if necessary and possible, but will keep them as long as possible',
+    not_idempotent = true,
+    input  = {'Lorem ipsum dolor sit amet, consec\004tetur adipiscing elit. Nunc vestibulum tempus ligula. Sed ac lobortis mi.'},
+    expect = '{\n    \'Lorem ipsum dolor sit amet, consec\\04\'...\n}',
+}
+
 --------------------------------------------------------------------------------
 -- Concatted Strings
 
@@ -167,6 +180,18 @@ format_test {
     expect = [['Lorem ipsum dolor sit amet, consectetur adipiscing elit.\004\002\000Nunc ve' ..]]..'\n'..[['stibulum tempus ligula. Sed ac lobortis mi.']],
 }
 
+format_test {
+    name   = 'Concatted string with decimal escape at border',
+    input  = 'Lorem ipsum dolor sit amet, consectetur adipiscing elit.Nunc vestibu\004\255\000lum tempus ligula. Sed ac lobortis mi.',
+    expect = [['Lorem ipsum dolor sit amet, consectetur adipiscing elit.Nunc vestibu\004' ..]]..'\n'..[['\255\000lum tempus ligula. Sed ac lobortis mi.']],
+}
+
+format_test {
+    name   = 'Concatted string with decimal escape at border 2',
+    input  = 'Lorem ipsum dolor sit amet, consectetur adipiscing elit.Nunc vestibu\004\002\000lum tempus ligula. Sed ac lobortis mi.',
+    expect = [['Lorem ipsum dolor sit amet, consectetur adipiscing elit.Nunc vestibu\004\02' ..]]..'\n'..[['\000lum tempus ligula. Sed ac lobortis mi.']],
+}
+
 
 --------------------------------------------------------------------------------
 -- Longform Strings

test/test_sorting.lua

@@ -247,3 +247,4 @@ end)
 --------------------------------------------------------------------------------
 
 return SUITE
+